Nickel Stabilization Efficiency of Aluminate and Ferrite Spinels and

Cover Image ... Adsorption and Thermal Stabilization of Pb and Cu by Zeolite ... Zinc Stabilization Efficiency of Aluminate Spinel Structure and its L...
0 downloads 0 Views 288KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 5520-5526

Nickel Stabilization Efficiency of Aluminate and Ferrite Spinels and Their Leaching Behavior KAIMIN SHIH,† TIM WHITE,‡ AND J A M E S O . L E C K I E * ,† Department of Civil and Environmental Engineering, Stanford University, Stanford California 94305-4020, and School of Materials Science and Engineering, Nanyang Technological University, Singapore 639798

Stabilization efficiencies of spinel-based construction ceramics incorporating simulated nickel-laden waste sludge were evaluated and the leaching behavior of products investigated. To simulate the process of immobilization, nickel oxide was mixed alternatively with γ-alumina, kaolinite, and hematite. These tailoring precursors are commonly used to prepare construction ceramics in the building industry. After sintering from 600 to 1480 °C at 3 h, the nickel aluminate spinel (NiAl2O4) and the nickel ferrite spinel (NiFe2O4) crystallized with the ferrite spinel formation commencing about 200-300 °C lower than for the aluminate spinel. All the precursors showed high nickel incorporation efficiencies when sintered at temperatures greater than 1250 °C. Prolonged leach tests (up to 26 days) of product phases were carried out using a pH 2.9 acetic acid solution, and the spinel products were invariably superior to nickel oxide for immobilization over longer leaching periods. The leaching behavior of NiAl2O4 was consistent with congruent dissolution without significant reprecipitation, but for NiFe2O4, ferric hydroxide precipitation was evident. The major leaching reaction of sintered kaolinitebased products was the dissolution of cristobalite rather than NiAl2O4. This study demonstrated the feasibility of transforming nickel-laden sludge into spinel phases with the use of readily available and inexpensive ceramic raw materials, and the successful reduction of metal mobility under acidic environments.

Introduction Hazardous heavy metals are among the most environmentally persistent wastes, are frequently highly mobile, and through leaching into water supplies can be bioaccumulated. Electroless nickel plating plays an important role in manufacturing, and the rapid increase of effluents arising from this process is of worldwide concern. Ocean discharge of plating wastes is prohibited under the London Dumping Convention, and regulations pertaining to such discharges are becoming increasingly stringent (1, 2). A survey revealed an average of 160 000 L/day of wastewater is discharged from a typical plating shop in the United States. With an estimated 3500 plating shops in the U.S., this represents about $1.45 billion treatment cost per year for the electroplating industry (3). * Corresponding author phone: 650-723-2524; fax: 650-725-3164; e-mail: [email protected]. † Stanford University. ‡ Nanyang Technological University. 5520

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 17, 2006

Physicochemical processes in use for heavy-metal removal from wastewater include precipitation, coagulation, reduction, ion exchange, and membrane processes. Among them, neutralization followed by sedimentation has been generally used in electroplating wastewater treatment because of low treatment cost and high treated water quality (4). However, this method results in the production of large amounts of heavy-metal bearing sludges which require additional treatment, such as immobilization, prior to landfill disposal. Many investigators have attempted to immobilize toxic metals via sorption, using natural or synthetic sorbents or cements, and then correlated performance directly with metal leaching (5-7). The immobilization processes where cementitious or pozzolanic materials are used have been received as an acceptable way to achieve solidificantion/ stabilization with reduced environmental risks associated with their subsequent disposal (8, 9). The cementation method uses the binder to either chemically bind toxic waste matter into solid bulk or physically cut them off from the outside by forming a capsule. For many types of wastes, including those from the industries of plating, anodizing, printed circuit board manufacture, and metal surface treatment, it is often referred as the best available and relatively inexpensive technology for pre-landfill treatment (10, 11). However, these solidification/stabilization technologies were not generally successful in preventing leaching in acidic environments, i.e., pH less than 4.0 (11, 12). In addition, the more stringent environmental requirements together with increasing operational and disposal costs have stimulated interest in possible alternative methods. While most metal-laden sludges are disposed of in controlled landfills, the option is expensive, encouraging many industries to seek alternative treatment options. Generators regulated by the U.S. EPA’s 40 CFR Part 261 “Identification and Listing of Hazardous Waste” (13) are attracted to permanent land-disposal alternatives due to increasing restrictions for the transportation and storage of sludge. In addition, metal wastes regulated under RCRA (Resource Conservation and Recovery Act) are required to meet rigorous standards prior to disposal in approved landfills. Subsequent amendments, such as the Land Disposal Restriction Regulation (LDR), banned the land disposal of untreated hazardous waste (14), and requires all treated wastes to be stored in secure landfills with no hydraulic contact, restricted access, and continuous monitoring of subsurface waters for contamination. This stimulated the development of other emerging technologies to reuse the waste and to replace the cementation method (8, 15). Radioactive waste has been successfully stabilized in vitreous and ceramic materials (16, 17), but these products are invariably contained in geologic repositories and cannot be considered for any reusable purpose. However, a similar thermal treatment, the lower firing temperatures normally used (900-1600 °C) in the brick, tile, and refractory aggregate industries may be helpful in promoting an effective incorporation of waste materials into the ceramic matrix (18, 19, 20), and also maintain the production of marketable materials. The purpose of this study is to reveal the stabilization efficiency and the product leaching behavior of a nickel stabilization strategy, which proposes to blend nickel sludge into the existing industrial sintering processes of construction ceramics. This strategy aims to produce safe and marketable metal-containing construction ceramics to avoid the sequential landfill disposal, and to reach virtually no or very little additional treatment cost to solve part of the hazardous metal sludge problem. Understanding of the temperature10.1021/es0601033 CCC: $33.50

 2006 American Chemical Society Published on Web 07/25/2006

dependent reaction mechanisms, the nickel incorporation efficiencies, the product phase leachability, and the leaching behavior are of great importance for evaluating the feasibility of this strategy. When thermally treated, most nickel-laden sludges decompose to nickel oxide (NiO), which is the most inert of common nickel compounds and remains stable until melting at around 1960 °C (21). Although NiO is relatively less soluble close to neutral pH ranges, excursions in acidity can lead to elevated nickel releases. Therefore, a search for more durable Ni-bearing phases is required. Our approach is to use kaolinite clay (Al2Si2O5(OH)4), a common ceramic precursor, that after calcination is converted to mullite and silica (22). Although readily available for fabrication of construction ceramics, kaolinite usually contains iron as a significant impurity, that upon heating separates as oxide(s), which have the potential to react with blended nickel. It has recently been found that nickel can be efficiently incorporated in nickel aluminate spinel (NiAl2O4) by sintering nickel oxide and kaolinite, via either a low-temperature reaction between nickel oxide and a defect spinel, or a hightemperature reaction between nickel oxide and mullite (18). NiAl2O4, which forms above 1350 °C (21), has been reported to have high resistance to acids and alkalis (19, 20). Nickel ferrite spinel (NiFe2O4), which also occurs as the mineral trevorite, has been reported in electric arc furnace dusts to be a low dissolution host for metals (23). However, the practicability of incorporating nickel into construction ceramics via spinel requires quantitative investigation of their chemical durability. In this study, the aluminum-rich precursors, kaolinite and γ-alumina (γ-Al2O3), and the iron-rich precursor, hematite (Fe2O3), were compared with respect to their nickel stabilization efficiency during sintering. In addition, the leaching behavior of the products was quantified to better assess the benefits of adopting this technology.

Experimental Methods Nickel oxide (NiO), kaolinite, γ-alumina (γ-Al2O3), and hematite were used as starting materials. NiO powder (Fisher Scientific) gave a measured surface area of 3.6 ( 0.5 m2/g after degassing by heating at 300 °C with He-gas purging for 3 h. The surface area was measured using a Beckman Coulter SA3100 surface area and pore size analyzer using the BET method. USP grade acid washed kaolinite powder from Fisher Scientific yielded a BET surface area of 9.0 ( 2.9 m2/g, and its powder X-ray diffraction (XRD) pattern matched kaolinite (ICDD PDF #78-1996). The diffraction patterns were collected using a Rigaku Geigerfex diffractometer (Rigaku Denki Co. Ltd.) equipped with a Cu X-ray tube operated at 35 kV and 15 mA. Scans were collected by a MDI XRD diffractometer control and data acquisition system (Material Data, Inc.) from 10 to 80° 2θ-angle, with the step size of 0.02° and a counting time of one second per step. The γ-Al2O3 was prepared from HiQ-7223 alumina powder (Alcoa Corporation) with a reported surface area of 228 m2/g and particle diameter of 54.8 µm in d50. As-received HiQ-7223 alumina was confirmed by XRD to be boehmite (AlOOH; ICDD PDF #74-1895), which after heat treatment at 975 °C for 3 h, transformed to a γ-Al2O3 dominant material (24, 25). The anhydrous iron (III) oxide powder (Fisher Scientific) was identified as hematite, Fe2O3. The XRD patterns of kaolinite, HiQ-7223 alumina, HiQ-7223 alumina heated at 975 °C for 3 h, and iron oxide powders are given in the Supporting Information. NiO and the tailoring precursors were mixed by ball-milling in a water slurry for 22 h, molded into 2 × 2 × 2 cm cubes and dried at 95 °C for 3 days. The dry cubes were homogenized by grinding, formed into 13 mm pellets at 125 MPa, and sintered at temperatures ranging from 600 to 1480 °C. The dwell time was fixed at 3 h to simulate short sintering typical of industrial production. The fired ceramics were air-

quenched and ground into powders for XRD analysis. In practice, construction ceramics will incorporate low levels of metal to ensure that product properties are not compromised. However, to study the incorporation efficiency of nickel-containing phases Ni:Al and Ni:Fe mole ratios of 1:2 were used, these being the stoichiometric maximum for nickel spinels. Nickel incorporation efficiency was monitored quantitatively by XRD. To determine the weight percent of each crystalline phase, diffraction patterns were processed using JADE (version 6.5.3) developed by Material Data, Inc. with a whole pattern fitting (WPF) function using the Pawley (26) method. A transformation ratio (TR) index was devised to monitor nickel incorporation efficiency into NiAl2O4 or NiFe2O4, such as the following:

transformation ratio (TR, %) ) wt% of spinel MW of spinel × 100 (1) wt% of spinel wt% of NiO + MW of spinel MW of NiO where MW is molecular weight. For TR ) 100%, complete transformation of nickel to spinel was achieved; for TR ) 0%, no nickel incorporation occurred. The application of this method was detailed in previous work (18), which reported (3% accuracy of TR values after screening the goodnessof-fit. To evaluate the leachability under prolonged acidic attack, the product phases were leached over 26 days using a pH 2.9 acetic acid solution, same as the extraction fluid no. 2 regulated under the U.S. EPA SW-846 method 1311 “Toxicity Characteristic Leaching Procedure (TCLP)”. The pH 2.9 acetic acid solution was prepared as the leaching agent from 5.7 mL glacial acetic acid and dilution with MilliQ water to a volume of 1 L. For the leach tests, cubes of NiO and precursors were fired directly at 1480 °C/48h, crushed into coarse particles, and then dry ball-milled for 8 h to ensure the homogeneous reduction of particle size. The surface areas of milled powder samples were measured using the BET method. Prolonged high-temperature sintering was used to ensure complete reaction and establish a baseline for comparison. Each leaching vial was filled with 10 mL extraction fluid and 0.5 g powder, and then tumbled endover-end at 60 rpm for periods of 0.75-26 days. At the end of each agitation, the leachates were filtered with 0.2-µm syringe filters, the pH measured, and the concentrations of Ni, Al, Fe, and Si derived with a TJA IRIS Advantage/1000 Radial ICP spectrometer calibrated against CLARITAS certified reference solutions (SPEX CertiPrep, Inc.).

Results and Discussion Nickel Incorporation Efficiency. Well-crystallized corundum (R-Al2O3) is formed after calcining γ-Al2O3 at high temperature. When NiO and γ-Al2O3 are sintered at 1250 °C, the product contains NiAl2O4, together with residual NiO and R-Al2O3 (Figure 1a), the overall reaction being

NiO + Al2O3 f NiAl2O4

(2)

The phase transformation of kaolinite under thermal treatment is known as the kaolinite-to-mullite reaction series (27, 28). During this process, physically bound water is removed at 110 °C, followed by decomposition of kaolinite with the loss of chemically bound water and transformation to amorphous metakaolin above 550 °C. Beyond 980 °C, metakaolin is converted to a poorly crystalline defect spinel phase. At 1000 °C, mullite starts to form with excess amorphous silica, which crystallizes as cristobalite upon further heating. It has recently been found that nickel reacts VOL. 40, NO. 17, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5521

FIGURE 1. XRD results of 1250 °C and 3 h sintering with the precursor of (a) NiO + γ-alumina, (b) NiO + kaolinite, and (c) NiO + hematite. “S” represents nickel aluminate spinel (NiAl2O4, ICDD PDF #78-0552); “N” for nickel oxide (NiO; ICDD PDF #78-0429); “A” for corundum (r-Al2O3; ICDD PDF #83-2080); “C” for cristobalite (SiO2; ICDD PDF #76-0938); “T” for nickel ferrite spinel (NiFe2O4, trevorite; ICDD PDF #86-2267). with both defect spinel and mullite to form NiAl2O4 during the sintering of kaolinite with nickel oxide (18), such that

NiO + Al8[Al13.3302.66]O32 or Si8[Al10.6705.33]O32 (defect spinel) 9 8 NiAl2O4 or NiAl2O4 + SiO2 (unbalanced) (3a) 980-1000 °C 3NiO + 3Al2O3‚2SiO2 (mullite) 9 8 >1000 °C 3NiAl2O4 + 2SiO2 (3b) The high-temperature reaction was confirmed by XRD (Figure 1b). Although iron oxide is nonessential for construction ceramics, it is present as a weight percent impurity in many precursors (29), including clays, silicates, and aluminates, and consequently, the reaction between iron oxide and NiO is of importance in waste stabilization. Hematite (Fe2O3) is a common oxide that results from thermal treatment of raw materials, and its reaction with nickel oxide can be written as follows:

NiO + Fe2O3 (hematite) f NiFe2O4 (trevorite)

(4)

and verified by XRD (Figure 1c). As noted earlier, sintering time on industrial production lines is limited, and optimizing incorporation efficiency as a function of sintering temperature essential. The transformation of NiO with γ-Al2O3, kaolinite or hematite to spinel as a function of temperature (600-1480 °C) at a fixed heating time of 3 h is shown in Figure 2. Nickel incorporation efficiency generally increased with temperature for both aluminum- and iron-rich precursors. For γ-Al2O3 samples sintered below 1100 °C, the presence of poorly crystalline transition alumina, which neither reacted with nickel to form NiAl2O4, nor transformed into corundum, led to high uncertainty in the quantitative estimation. Similarly when kaolinite was the precursor, WPF analysis could only consider the crystalline phases (NiO, NiAl2O4, mullite, and cristobalite), and phase quantification was not possible due to the presence of amorphous silica. However, from analysis of all the available data, it can be estimated that the formation temperature of NiFe2O4 is about 200-300 °C lower than for 5522

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 17, 2006

FIGURE 2. Nickel incorporation efficiency with γ-alumina, kaolinite, or hematite precursor. The values of transformation ratio (TR) were plotted for NiO + γ-alumina samples, NiO + kaolinite samples, and NiO + hematite samples with Ni/Al or Ni/Fe mole ratio of 1:2, and sintered at corresponding temperatures for 3 h. Error bars were plotted by data ranges from some experiments with three independent replicates. NiAl2O4 from both the γ-Al2O3 or kaolinite precursors. Consequently, it is expected that iron-rich spinels will crystallize first when sintering of nickel-laden sludge with precursors enriched in both aluminum and iron. WPF suggests a small amount of unreacted NiO remains even after high-temperature sintering, and the maximum TR value was ∼98%. However, quantification was difficult, because all the reflections of NiO (2θ ) 37.25°, 43.29°, 37.25°, 62.88°, 75.41°, 79.41°) overlay those of NiFe2O4, possibly leading to a systematic error. Nevertheless, the TR values reflect very high nickel incorporation efficiency. Leachability. The leaching experiment described earlier was used to evaluate the prolonged leachability of NiO, NiAl2O4, NiFe2O4, and the product derived from NiO + kaolinite. Although theoretical solubility constants of some of these solid phases at equilibrium are reported, an understanding of leachability and dissolution kinetics under regulatory conditions is required. NiO is known to be sparingly soluble and NiAl2O4 is reported as highly resistant to acids and alkalis, but to further differentiate the leachability of these phases, extended testing is required. As the current focus is on pure phase comparisons, the NiO + γ-Al2O3, NiO + Fe2O3, and NiO + kaolinite samples sintered at 1480 °C for 48 h were selected for evaluation where the TR values derived from XRD were ∼100%. Figure 3 shows the scanning electron micrographs of polished sintered products where backscattered electrons give rise to contrast that is proportional to average atomic number. These micrographs reveal homogeneous spinel-phases in sintered NiO + γ-Al2O3 and NiO + Fe2O3 samples, with the latter being more porous. On the other hand, NiO + kaolinite sintered under the same conditions separated into aluminum-rich (NiAl2O4) and silicon-rich (cristobalite) areas (Figure 3c). These samples were ground into powders and measured for BET surface area to yield values of 1.1 ( 0.1 m2/g for NiAl2O4, 1.7 ( 0.2 m2/g for NiFe2O4, and 0.73 ( 0.12 m2/g for sintered NiO + kaolinite. During leach testing, thirteen aliquots were taken over a 0.75-26 days period. Together with NiO, three independent replicates were made for each experiment.

FIGURE 3. Scanning electron micrographs of polished surface of 1480 °C/48 h sintered (a) NiO + γ-Al2O3; (b) NiO + Fe2O3; (c) NiO + kaolinite. Sample (a) and (b) showed homogeneous spinel phases (NiAl2O4 and NiFe2O4 respectively), except some dark grain boundaries in (a) and voids in (b). The light-color grain in sample (c) showed the enrichment of Ni and Al (NiAl2O4), and the darker matrix showed the enrichment of Si (cristobalite). The pH of the leachates are shown in Figure 4. In the first few days there is a rapid increase in pH, but after 5 days the spinel leachates stabilized, while NiO leachate pH increased for the remainder of the experiment. The increase in pH arose from the dissolution of cations that exchange with protons in solution, accompanied by the destruction of crystals. It is apparent that NiO is more vulnerable to protonmediated dissolution. Figure 4 also shows relatively higher pH for NiAl2O4 leachates, as compared to the other two nickelspinel containing samples (NiFe2O4 spinels and sintered NiO + kaolinite samples). This could imply higher nickel leachability of NiAl2O4 compared to NiFe2O4; on the other hand, the lower leachate pHs of sintered NiO + kaolinite samples (NiAl2O4 + cristobalite) provides contradictory evidence. Figure 5 summarized the leachate nickel concentrations normalized with respect to nickel content and surface area of powder samples. Long-term leaching will be dominated by surface reactions, and is expected to be proportional to surface area. In addition, since the same weight of sample (0.5 g) was always used, the total nickel content in the sample, subject to the different nickel phases, should also be normalized for comparison. After the first 2-3 days of leaching, the difference in nickel concentration among the sample leachates was indistinguishable. However, over longer periods all spinel samples proved to be superior stabilization matrixes compared to NiO, although NiFe2O4 may be a slightly more leachable spinel than NiAl2O4. Elevated short-term leachability may be due to incomplete spinel formation or acid attack at grain boundaries. While the nickel concentration of the NiAl2O4 and NiFe2O4 leachates are similar, the substantial difference in pH implies the leaching mechanisms

might be different. In the case of sintered NiO + kaolinite samples, it should be noted that much lower nickel releases were observed even though the sample surface area and nickel content were proportionally normalized. These experiments again demonstrate that the prescribed short leaching time (18 h) of the TCLP is incapable of distinguishing the relative leachabilities of the product phases nor can it predict the long-term stability of ecomaterials. Leaching Behavior. The pH and nickel concentration data reveal two behavioral differences: (1) the greater acidity of NiFe2O4 leachates compared to NiAl2O4, and (2) the lower nickel leachability of sintered NiO + kaolinite samples as compared to NiAl2O4. To further investigate their leaching behavior the concentration of the other metal ions were considered. In Figure 6, the [Ni]/[Al] and [Ni]/[Fe] leachate molar ratios of NiAl2O4 and NiFe2O4 are shown. Congruent dissolution of spinels without reprecipitation should result in leachates with [Ni]/[Al] or [Ni]/[Fe] molar ratios of 0.5, and for NiAl2O4 the [Ni]/[Al] molar ratio was maintained around 0.48-0.50 (Figure 6a). However, in the case of NiFe2O4, the [Ni]/[Fe] molar ratio varied from 30 to 50 after the first few days. This indicates significant reprecipitation of iron and/or the nickel was incongruently dissolved from NiFe2O4 (Figure 6b). The iron concentration in NiFe2O4 leachates remained steady at 0.5-1.0 ppm (Figure 6c), a result consistent with the solubility constant (30) of amorphous ferric hydroxide (am‚Fe(OH)3(s)) at pH 3.0 (according to the measured pH values in Figure 4) and ionic strength around 1.5 × 10-3M of the system. The calculation details are provided in the Supporting Information. This estimated FeT value (0.90 ppm) VOL. 40, NO. 17, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5523

FIGURE 4. pH values of the sampled leachates of NiO, 1480 °C/48 h sintered NiO + γ-Al2O3 sample (NiAl2O4), 1480 °C/48 h sintered NiO + kaolinite sample (NiAl2O4 + cristobalite), and 1480 °C/48 h sintered NiO + hematite sample (NiFe2O4). The leaching solution was TCLP extraction fluid no. 2 (acetic acid solution) at pH 2.9. Each leaching vial was filled with 10 mL extraction fluid and 0.5 g powder, and then tumbled end-over-end at 60 rpm between 0.75 and 26 days.

FIGURE 6. [Ni]/[Al] molar ratios of NiAl2O4 spinel leachates (a), [Ni]/[Fe] molar ratios of NiFe2O4 spinel leachates (b), and soluble iron concentration in the leachates of leaching NiFe2O4 spinel (c) by pH 2.9 TCLP extraction fluid no. 2 (acetic acid solution) tumbled end-over-end at 60 rpm between 0.75 and 26 days. The ratios of solid to fluid weight were 1:20, and the solids were in a Ni/Al or Ni/Fe mole ratio of 1:2. as a result of the production of hydronium ions during precipitation:

FIGURE 5. Normalized nickel concentrations of the leachates of NiO, 1480 °C/48 h sintered NiO + γ-Al2O3 sample (NiAl2O4), 1480 °C/48 h sintered NiO + kaolinite sample (NiAl2O4 + cristobalite), and 1480 °C/48 h sintered NiO + hematite sample (NiFe2O4). The surface area of NiO powder was 3.6 ( 0.5 m2/g, NiAl2O4 powder was 1.1 ( 0.1 m2/g, sintered NiO + kaolinite powder was 0.73 ( 0.12 m2/g, NiFe2O4 powder was 1.7 ( 0.2 m2/g. The leaching solution was TCLP extraction fluid no. 2 (acetic acid solution) at pH 2.9. Each leaching vial was filled with 10 mL extraction fluid and 0.5 g powder, and then tumbled end-over-end at 60 rpm between 0.75 and 26 days. was close to the observed iron concentrations shown in Figure 6c. Therefore, it is suggested that iron released during the leaching of NiFe2O4 by TCLP extraction fluid no. 2, and the iron concentrations of leachates were controlled through equilibrium with amorphous ferric hydroxide as the solid phase. Dissolution followed by ferric hydroxide precipitation can also explain the lower pH values of NiFe2O4 leachates, 5524

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 17, 2006

NiFe2O4 + 8H+ f Ni2+ + 2Fe3+ + 4H2O

(8)

Fe3+ + 3H2O f am‚Fe(OH)3(s) + 3H+

(9)

To better understand the leaching behavior of sintered NiO + kaolinite samples, the [Ni] and [Si] molar concentrations are shown in Figure 7. From electron microscopic images, the microstructure of sintered NiO + kaolinite consists of NiAl2O4 crystals embedded in a cristobalite (SiO2) matrix. The mole ratios of Ni/Si in powder samples were 1:2. If sintered NiO + kaolinite samples leached congruently without reprecipitation, the [Ni]/[Si] molar ratios of leachates should be around 0.5 throughout the leaching period. However, as summarized in Figure 7, the ratios are much lower and with silicon concentrations increasing continuously. This indicates the major leaching reaction of sintered NiO + kaolinite samples was the dissolution of cristobalite, rather than leaching of NiAl2O4 spinel. The reason the leachability of NiAl2O4 in sintered NiO + kaolinite sample was lower than leaching NiAl2O4 spinel alone (sintered NiO + γ-Al2O3 samples in Figure 5) is possibly due to enhanced spinel crystallization

Literature Cited

FIGURE 7. Nickel and silicon concentrations in the leachates of leaching sintered NiO + kaolinite samples by pH 2.9 TCLP extraction fluid no. 2 (acetic acid solution) tumbled end-over-end at 60 rpm between 0.75 and 26 days. The ratios of solid to fluid weight were 1:20, and the solid samples were in a Ni/Si mole ratio of 1:2. and robust grain boundaries promoted by the silica flux. Most commercial construction ceramics have silicon oxide as a major component, and will be similarly vulnerable under acid attack. Although the destruction of the SiO2 matrix will decrease the physical strength of the ceramic, most incorporated nickel will be crystallochemically stabilized in the more durable NiAl2O4 spinel. Since sintering NiO + kaolinite is the more practical strategy in making low-cost nickelcontaining ceramics, this leaching behavior would prove to be an additional benefit in stabilizing nickel from a kaolinitebased precursor. Future studies might include using Rietveld analysis to quantify the amorphous phase that appears at lower sintering temperatures, additional studies of leaching performance of NiAl2O4-SiO2 composites, and evaluating the possible role of the NiAl2O4-NiFe2O4 solid solution in systems enriched simultaneously with iron and aluminum.

Acknowledgments Special thanks are due Professor J. F. Stebbins for the use of equipment in pelletizing and sintering, to Mr. R. E. Jones for the assistance on XRD analysis and scanning electron microscopy, and to Dr. G. Li for the ICP spectrometer metal measurements. The Alcoa Corporation is acknowledged for providing HiQ-7223 alumina. We also like to thank Professor Z. Dong and Professor M. Reinhard for fruitful discussions. Support from the Singapore Clean Water Programme and the Singapore Stanford Partnership is acknowledged.

Note Added after ASAP Publication There was an error in equation 3a in the version published ASAP July 25, 2006; the corrected version was published ASAP July 31, 2006.

Supporting Information Available One Figure and one table showing the XRD patterns of kaolinite powder, HiQ-7223 alumina powder, 975 °C and 3 h treated HiQ-7223 alumina, iron oxide powder, and the dissolution of iron hydroxide at pH 3.0 with ionic strength of 1.5 × 10-3M. This material is available free of charge via the Internet at http://pubs.acs.org.

(1) International Maritime Organization. The London Dumping Convention; International Maritime Organization: London, 1991. (2) Horikawa, K.; Hirasawa, I. Removal and recovery of nickel ion from wastewater of electroless plating by reduction crystallization. Korean J. Chem. Eng. 2000, 17, 629-632. (3) USEPA. Nickel Plating: Industry practices control technologies and environmental management, USEPA Capsule Report 635/ R-03/005; U.S. Government Printing Office: Washington, DC, 2003. (4) Park, D.; Lee, D. S.; Park, J. M.; Chun, H. D.; Park, S. K.; Jitsuhara, I.; Miki, O.; Kato, T. Metal recovery from electroplating wastewater using acidophilic iron oxidizing bacteria: Pilot-scale feasibility test. Ind. Eng. Chem. Res. 2005, 44, 1854-1859. (5) Kapoor, A.; Viraraghavan, T. Treatment of metal industrial wastewater by fly ash and cement fixation, J. Environ. Eng. ASCE. 1996, 122, 243-243. (6) Lin, C.-F.; Lo, S.-S.; Lin, H.-Y.; Lee, Y. Stabilization of cadmium contaminated soils using synthesized zeolite. J. Hazard. Mater. 1998, 60, 217-226. (7) Bailey, S. E.; Olin, T. J.; Bricka, R. M.; Adrian, D. D. A review of potentially low-cost sorbents for heavy metals. Water Res. 1999, 33, 2469-2479. (8) Rha, C. Y.; Kang, S. K.; Kim, C. E. Investigation of the stability of harden slag paste for stabilization/solidification of wastes containing heavy metal ions. J. Hazard. Mater. 2000, B73, 255267. (9) Viguri, J.; Andre´s, A.; Ruiz, C.; Irabien, A.; Castro, F. Cementwaste and clay-waste derived products from metal hydroxides wastessEnvironmental characterization. Process Saf. Environ. Prot. 2001, 79(B1), 38-44. (10) Cheeseman, C. R.; Butcher, E. J., Sollars, C. J.; Perry, R. Heavy metal leaching from hydroxide sulphide and silicate stabilized/ solidified wastes. Waste Manage. 1993, 13(8), 545-552. (11) Yousuf, M.; Mollah, A.; Vempati, R.; Lin, T.; Cocke, D. The interfacial chemistry of solidification/stabilization of metals in cement and pozzolanic material systems. Waste Manage. 1995, 15, 137-148. (12) Bonen, D.; Sarkar, S. L. The effect of simulated environmental attack on immobilization of heavy metal doped in cementbased materials. J. Hazard. Mater. 1995, 40, 321-335. (13) Office of the Federal Register. Identification and listing of hazardous waste, Code of Federal Regulations Title 40, Part 261; U.S. Government Printing Office: Washington, DC, 1998. (14) Bostick, W. D.; Hermes, W. H. Control of regulated metals at DOE Oak Ridge reservation sites, Waste Manage. 1993, 13, 439454. (15) Yang, G. C. C.; Kao, K. L. Further study on properties of a solidified electroplating sludge using a mixture of an electroplating sludge and a calcium carbonate sludge as a binder, J. Hazard. Mater. 1994, 39(3), 301-315. (16) Lewis, M. A.; Fischer, D. F.; Smith, L. J. Salt-occluded zeolite as an immobilization matrix for chloride waste salt, J. Am. Ceram. Soc. 1993, 76, 2826-2832. (17) Lewis, M. A.; Fischer, D. F.; Murphy, C. D. Properties of glassbonded zeolite monlith. Ceram. Trans. 1994, 45, 277-286. (18) Shih, K.; White, T.; Leckie, J. O. Sintering of simulated Ni-bearing sludge and kaolinite: Ni-Al spinel formation. Proceedings, Joint Clean Water Programme Symposium; Singapore, 2005. (19) Teixeira da Silva, V. L. S.; Lima, F. P.; Dieguez, L. C.; Schmal, M. Regeneration of a deactivated hydrotreating catalyst. Ind. Eng. Chem. Res. 1998, 37, 882-886. (20) Cesteros, Y.; Salagre, P.; Medina, F.; Sueiras, J. E. Synthesis and characterization of several Ni/NiAl2O4 catalysis active for the 1,2,4-trichlorobenzene hydrodechlorination. Appl. Catal., B 2000, 25, 213-227. (21) Phillips, B.; Hutta, J. J.; Warsaw, I. Phase equilibria in the system NiO-Al2O3-SiO2. J. Am. Ceram. Soc. 1963, 46, 579-583. (22) Murray, H. Industrial Clays Cases Study; International Institute for Environment and Development: England, 2002. (23) Stegemann, J. A.; Roy, A.; Caldwell, R. J.; Schilling, P.; Tittsworth, R. Understanding environmental leachability of electric arc furnace dust. J. Environ. Eng. 2000, 126, 112-120. (24) Zhou, R.-S.; Snyder, R. L. Structures and transformation mechanisms of η, γ and θ transition aluminas. Acta Crystallogr. 1991, B47, 617-630. (25) Wang, Y.; Suryanarayana, C.; An, L. Phase transformation in nanometer-sized γ-alumina by mechanical milling. J. Am. Ceram. Soc. 2005, 88, 780-783. (26) Pawley, G. S. Unit-cell refinement from powder diffraction scans. J. Appl. Crystallogr. 1981, 14, 357-361. VOL. 40, NO. 17, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5525

(27) Brindley, G. W.; Nakahira, M. The kaolinite-mullite reaction series: I, A survey of outstanding problems. J. Amer. Ceram. Soc. 1959, 42, 311-314. (28) Brindley, G. W.; Nakahira M. The kaolinite-mullite reaction series: II, Metakaolin. J. Am. Ceram. Soc. 1959, 42, 314-318. (29) Sun, D. D.; Chang, L.; Tay, J. H.; Navratil, J. D.; Easton, C. Recovery and marine clay stabilization of heavy metals present in spent hydrotreating catalysts. J. Environ. Eng., ACSE 2001, 127, 916921.

5526

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 17, 2006

(30) Morel, F. M. M.; Hering, J. G. Principles and Applications of Aquatic Chemistry; John Wiley and Sons: New York, 1993. (31) Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; John Wiley and Sons: New York, 1996.

Received for review January 17, 2006. Revised manuscript received June 3, 2006. Accepted June 5, 2006. ES0601033